A Study On The Impact Of Social Media On The Volatility Of Chinese Stock Market

Financial market volatility forecasting has always been one of the hot topics in the financial field,and realized volatility,as a representative volatility,reflects investors’attitudes and perceptions of the market.Sentiment plays an important role in investors’investment decisions and has an impact on market volatility.Therefore,by studying the relationship between social media sentiment and realized volatility,it provides implications for investors to make decisions and market managers to formulate policies to deal with financial market volatility.In this paper,we examine the impact of social media sentiment on market volatility and stock volatility in general,positive,negative and neutral,and compare and analyze the predictive power of social media sentiment on realized volatility models in the market and in the CSI 300 stocks.First,this paper will select the total post volume,positive post volume,negative post volume and neutral post volume of Eastern Fortune stock bar as social media sentiment indicators.Second,this paper selects HAR-RV,SHAR-RV and HAR-CJ models as benchmark models for predicting realized volatility of SSE index and realized volatility of each constituent of CSI 300 index,and adds social media sentiment to the benchmark models to test the prediction differences on volatility.In addition,this paper investigates the predictive power of social media sentiment on future 1-day,2-day,1-week,2-week,and 1-month volatilities,and then evaluates the model in a comprehensive manner using the out-of-sample!!~#!"statistic,the DM test,and the MCS confidence level.Finally,the robustness of the findings is verified using a new windowing technique,rolling with different window lengths and grouping of high and low volatilities.The main conclusions of this paper are(1)Social media sentiment has the risk of enhancing stock market volatility and the asymmetric effect is significant.(2)The impact of social media sentiment on stock market volatility has a time-sensitive feature.(3)Social media sentiment has a dispositional effect on individual stock volatility.(4)Social media sentiment on high and low volatility on the market and individual stock performance is different.Combined with the findings of this paper,it is recommended that individual investors be cautious in making decisions based on social sentiment;it is recommended that institutional investors pay attention to positive social media sentiment and analyze more from outside social media when the market fluctuates significantly;it is recommended that regulation strengthen investor opinion analysis to avoid extreme fluctuations in social media sentiment.